I love coffee.
I’ve always loved the smell of coffee, but I haven’t always liked how coffee tastes. There was always a disappointing disconnect between the two flavors — I drink coffee for business and I smell it for pleasure.
This all changed in February 2011 when I was on a trip to New York with Mathieu, my manager at Google. He took me to a cafe in Chelsea and bought me a cappuccino. The bearded barista took his time lovingly making the espresso, texturing and foaming the milk, and pouring a broad-leafed rosetta into a warm cup. The result was delicious, and was the closest any drink had come to rendering the aroma of coffee as a liquid that I had ever had — a far cry from the brown wake-up juice I was used to.
I was an instant addict. I went back to Boston and found that I was indignant at cups of coffee that I would have been satisfied with before. What’s more, I was completely disconnected from Boston’s active coffee scene. In one interaction with a cafe owner typical of that period, after receiving an exceptionally bitter cup of coffee, I insisted that a cafe latte should be perfectly smooth, without a hint of bitterness. He said, “Sir, that’s how coffee tastes. If you would like a drink without bitterness, I recommend a drink with caramel syrup.” I responded, “You overextracted the coffee and boiled the milk. If that’s the way you like your coffee, that’s your prerogative as a barista. My prerogative as a customer is to never come back to this cafe again,” and I left.
Eventually I decided I would learn to make coffee myself. Nearly four years later, I still feel like I am at the beginning of understanding the gap between the two flavors. I’ve met a lot of passionate people who love coffee, who taught me a lot and shared many cups of coffee prepared in many different ways. Somehow coffee preparation is still a mysterious enterprise full of rituals that may or may not have an impact on the resulting drink. There are dozens of variables to control, and flavor, aroma, and texture are subjective experiences that are difficult to quantify.
Here’s what I believe about coffee today.
Coffee drinks are a suspension and a solution of molecules that are found in roasted coffee beans in water. When green beans are roasted, they undergo reactions that create molecules that are responsible for a lot of the aromas, colors and flavors of the resulting drink. The main reaction is called the Maillard reaction — it’s one of the two chief non-enzymatic browning reactions in flavor science (the other being caramelization). Roasted beans contain chemicals that having varying levels of solubility in water. Certain chemicals, such as acids, sugars, caffeine, and tannins will readily dissolve in water; on the other hand, oils and other hydrophobic compounds will not dissolve in water.
There are many different methods of preparing coffee; they can be classified by how they deal with low solubility compounds. This is not a judgment, by the way; these different methods (with the possible exception of the first) can be done well or done poorly and yield drinks that are not necessarily better or worse than each other, but are simply different.
- They can be ignored (instant coffee, very water soluble).
- Accept the low extraction rate, wait a long time (cold brew).
- Increase the extraction rate with high temperature (drip coffee).
- Increase the extraction rate with high pressure (espresso).
- Increase the extraction rate with high agitation (does anyone do this?).
This may not be all there is to it. It seems likely that brewing conditions will cause chemical reactions (assembling or breaking down molecules) beyond merely physically extracting them from the bean, but I don’t know anything about this — it’s only a guess.
Likewise, I have no idea what the chemicals in the beans actually are. I know there are thousands of them, but I can only name a handful.
It seems generally accepted that high solubility compounds are watery and sour and the low solubility compounds are oily and bitter. Many of the variables that affect texture and flavor affect extraction in a simple way. For example, grinding the beans more finely increases extraction rates (since the water will have greater contact with the bean), which will make the drink less sour and more bitter. Similarly, increasing dwell time (the amount of time the grinds are in contact with the water) will extract more low-solubility compounds. Almost every variable that could conceivably affect flavor and texture does so, from how the coffee is roasted and ground all the way to the humidity at the moment it is prepared.
The resulting beverage can be extraordinarily sensitive to these variables. Most notably, a difference in the temperature of the brewing water of less than 1°F (0.5°C) can be the difference between a sour cup and a sweet cup, or between a sweet cup and a bitter cup. Traditional coffee equipment does not have mechanisms for precise temperature control; modern high-end equipment will use tuned control loops (commonly PID) to precisely control the temperature of brewing water.
Experienced baristas are aware that these variables affect flavor, but they may or may not have access to measurement equipment or brewing equipment that gives them enough control to experiment.
Coffee is such a complex beverage; it’s clear we’re at the tip of the iceberg. I’m hoping we can design equipment that is inexpensive enough to give more control of more variables to more people.
More on how we will do this soon.